Ogre Forums

volca

Damn it does look fine, even tried the demo and will be strongly tempted to buy it when it comes out. I really wanted to push openDarkEngine all the way to complete game, but I lost the wind to continue because the original source code leaked and is probably the reason we see unofficial patches that...

volca

Woot, got the uv coords working correctly. http://i.imgur.com/FAJrs2f.jpg Next up is objects. The levels are a bit bare without props. I'm going to pass on light maps for now. They are very low res and only 15 bit colour. I'm going to try porting to Ogre 2.1 and see how it's lighting/shadows make i...

volca

You know, this was just a cold blooded experiment if someone replies when I write something off the charts. Something has surely changed because this was not the behavior of ogre 1.4 or 1.6 (dunno about the lattter). I'll look into the problem you mention. Don't you think this should all go into som...

volca

Is ogre essentially a dead or undermanned project now? I have to admit I am quite unhappy with the way the patch was rejected. https://bitbucket.org/sinbad/ogre/pull- ... h/activity

volca

Hey,

I am really looking forward to see this in vanilla ogre - are there any plans to get this into v2-0 or is it not a viable time frame to fill?

volca

Looking at v2-0 the same problem is there in OgreZip.cpp, I'd be delighted if this bug wasn't there, at least in v2-0. The doc still says: /// The file's fully qualified name String filename; /// Base filename String basename; There is no reason to have a duplicit info in the FileInfo struct, AMIRITE?

volca

I've posted a pull request on the zip archive issue and will look at the cmake scripts soon.

volca

Thanks for your reply, I'll report both issues as soon as I get the time to do it.

In the meantime I'll probably just clone and fix the OgreZip.cpp/.h anyway, so patch is not a problem.

volca

Okay, looking at the source code, there seems to be an extra branch in OgreZip.cpp load method: + else + { + info.filename = info.basename; + } This seems to have been removed in changeset 3566:047d710287b5, and got back in 4740:0adb7d41cba9 Is there any agreement on how the archive should work? fil...

volca

Hey Guys, yesterday I spent my evening porting my project from ogre 1.7 to ogre 1.9 and encountered a few troubling bugs. So I'd like you to tell me I'm doing something wrong :) 1. OGRE.pc on Arch prefix=/usr exec_prefix=${prefix} libdir=${prefix}/lib includedir=${prefix}/include plugindir=${libdir}...

volca

Okay, sorry for the false alarm :) I still find it alarming that the self-copy construction seems to be valid (g++ didn't even spit a warning when I tried it). EDIT: At least clang identifies the problem: a.cc:11:8: warning: variable 'a' is uninitialized when used within its own initialization [-Wun...

volca

Hey,

just browsing through the commits - is this intentional?
https://bitbucket.org/sinbad/ogre/commi ... de/Grass.h

in think the vMax construction is wrong. Perhaps change:

Code: Select all

Vector3 vMax( -vMax );

to

Code: Select all

Vector3 vMax( -vMin );

volca

While the site does not seem to contain the problem mentioned I suppose it does not hurt to mention it: http://graphics.stanford.edu/~seander/bithacks.html

volca

Well you can still package that and forward as the first parameter to the std::function.

Code: Select all

struct Callback {
    std::function<void(Callback*, /*other params*/)> callback;
    std::string name;
    // etc.
};

volca

Personally I'd rework the console to use std::function instead and then use std::bind.

volca

Well you could do a helper function for this, skipping the need for a Vector3 method:

Code: Select all

inline void trucate(Ogre::Vector3 &v, double maxLength) {
  if(v.squaredLength() > (maxLength * maxLength)) {
    v = (v.normalisedCopy() * maxLength);
  }
}

Code: Select all

truncate(v, maxLen);

volca

I'd be worried a bit about the fact that it is hard to control how java does garbage collection and memory management in general. JIT compiler will probably eat some resources as well (and that is not easily controllable either). I think adapting existing scripting language to LLVM could be a viable...

volca

I think the propper way to resolve this is to transition to 0x and use unique_ptr which does not have the hacky ownership handing c'tor and uses move semantics instead.

Edit: Oh, late to the discussion - already mentioned

volca

You could try doing a coredump and analyze that instead:

Code: Select all

ulimit -c unlimited
<run ogre meshy...>
gdb -c core <ogre_meshy_binary>

volca

If you'd be working under linux, I'd tell you to try using valgrind - chances are it could find a problem (don't know about such a thing for windows). Other ideas - you're modifying the maps content whilst iterating it (in a such way it relocates/modifies the current element), the object which held ...

volca

Why don't you use something like tbb::concurrent_queue? You don't have to worry about concurrency then.

volca

If you'd be still wanting to access the fields of the structs, you'd have to supply that info to the code working with the structs. I once written something similar - There was a function that was called at the beginning of the code execution (once) and mapped the fields of the struct one by one (th...

volca

After the textures are atlased how difficult is it to modify the texture co-ords of the mesh to match up? This depends on your meshes - if you have the meshes unwrapped (that is if the texturing coordinates never go out of 0-1.0f range), it is a simple thing to do. If not, this is not entirely triv...

volca

I think this is best done offline - there are probably too many "Ifs" in my opinion - the materials have to be similar (how do I compare?), the geometry has to be unwrapped, etc. The implementation itself would not be that hard to do though (meaning if your meshes/materials are prepared properly, th...

volca

Sorry for the delay. The code fails in + operator of basic_string (That is used when composing the exception text). The reason is that std::vector::operator[] has undefined behavior for index out of the range 0-size() - so it gives back an invalid std::basic_string reference - in my case it is Null ...

Ogre Forums

Search found 390 matches

Re: Ogre WIP and random screenshot thread (for everyone!)

Re: Ogre WIP and random screenshot thread (for everyone!)

Re: A few issues

Re: A few issues

Re: [GSoC 2013 - accepted] Resource System Redesign

Re: A few issues

Re: A few issues

Re: A few issues

Re: A few issues

A few issues

Re: [GSoC 2013 - accepted] Ogre 2.0

Re: [GSoC 2013 - accepted] Ogre 2.0

Re: Endianness and Bitwise Operators

Re: Console design problem

Re: Console design problem

Re: Any support for a 'truncate' method within Vector?

Re: Bohemia Interactive (OFP, ARMA) will use Java for script

Re: My dark auto_ptr<>

Re: Ogre Meshy 1.21 - (Mesh viewer, Now supports Linux!)

Re: Problem with a template function?

Re: Accessing a vector of structs in a new thread

Re: Sending a struct to a class

Re: Atlasing addon for StaticGeometry?

Re: Atlasing addon for StaticGeometry?

Re: Small bug in TextureUnitState::getTextureDimensions